Peter Geoghegan <peter(at)2ndquadrant(dot)com> writes:
> Why does commit_delay have to be an integer? Can't we devise a way of
> manipulating it in units of milliseconds, but have the internal
> representation be a double, as with pg_stat_statements' total_time
> column?

If you wanted to re-implement all the guc.c logic for supporting
unit-ified values such that it would also work with floats, we could
do that.  It seems like way more mechanism than the problem is worth
however.

			regards, tom lane

In response to

• Re: [COMMITTERS] pgsql: Revert "commit_delay" change; just add comment that we don't hav at 2012-08-15 10:51:54 from Peter Geoghegan

Responses

• Re: [COMMITTERS] pgsql: Revert "commit_delay" change; just add comment that we don't hav at 2012-08-15 15:41:38 from Peter Geoghegan

pgsql-hackers by date

Next:	From: Tom Lane	Date: 2012-08-15 13:48:43
	Subject: Re: sha1, sha2 functions into core?
Previous:	From: Marko Kreen	Date: 2012-08-15 11:42:13
	Subject: Re: sha1, sha2 functions into core?

pgsql-committers by date

Next:	From: Bruce Momjian	Date: 2012-08-15 15:23:17
	Subject: pgsql: Properly escape usernames in initdb,so names with single-quotes
Previous:	From: Peter Geoghegan	Date: 2012-08-15 10:51:54
	Subject: Re: [COMMITTERS] pgsql: Revert "commit_delay" change; just add comment that we don't hav